Adenocarcinoma of the pancreas is a highly virulent neoplasm with a dismal prognosis. In 1991, it was estimated that 28,200 new cases will be diagnosed, and over 25,000 patients will die of this disease. The mean survival of patients with pancreatic cancer remains in the 4 to 6 month range, with long-term survival rates showing less than 5% of patients can be expected to survive over 5 years. The most common symptoms seen in pancreatic cancer are jaundice, anorexia, weight loss, and pain. Estimates show that 70 to 90% of patients with pancreatic cancer will experience pain at some time during the course of their illness. Thus, the palliation of pain in patients with pancreatic carcinoma remains one of the most important aspects of the overall management of this disease. Currently there are two acceptable and advocated methods for palliation of pancreatic carcinoma pain. These are: external beam radiation therapy and percutaneous celiac axis chemical splanchnicectomy. Despite the multiple reports on both of these topics, there is little objective data with respect to the ability for these treatments to adequately palliate pancreatic cancer pain. In conjunction with the Memorial Sloan-Kettering Cancer Center we plan a trial which specifically plans to prospectively evaluate, using a random assignment study, the efficacy of radiation therapy or of celiac axis block in relieving pain and psychological distress in patients with advanced, unresectable adenocarcinoma of the pancreas. We hope to objectively define the overall time course of these therapies, both in respect to onset and duration of pain relief and their complications and/or side effects. We also hope to evaluate the quality of life in patients undergoing these treatments. As secondary goals, we hope to evaluate survival in this patient population and to determine the relative financial costs of the two treatment modalities. All patients will undergo a battery of standard, validated tests of pain, mood, and quality of life prior to randomization. Only patients with severe pain, unresponsive to standard oral opiate pain management will be randomized. The patient's pain will be assessed throughout treatment and in follow-up until death.